INFINITIVE · STORYTELLING

INSTITUTIONAL SIGNAL AUDIT — PILOT

SUPERCITY

CIVIC-AI · CITY OF EL PASO

Where the signal breaks before the system does.

BASIS Public records + first-party capture CONFIDENCE CEILING SUPPORTED DATE June 18, 2026

01 WHAT THIS IS METHOD

An Institutional Signal Audit is a one-week, confidence-rated read of where an institution, program, or AI deployment is most exposed to public governance failure — before it breaks. It does not adjudicate the underlying technology or anyone's intent. It examines the gap between the signal an institution projects and the signal its public actually receives, and names the single exposure most worth closing first.

This pilot was run against SuperCity's public footprint as a demonstration of method. It is offered to the City and to SuperCity in a constructive spirit: the finding below is common to fast civic-AI deployments, it is fixable, and fixing it strengthens the partnership rather than threatens it.

On confidence. Every claim carries a rating — ESTABLISHED (primary evidence, survives adversarial review), SUPPORTED (consistent evidence, alternatives not excluded), PLAUSIBLE, SPECULATIVE. What this audit cannot establish from public information is stated plainly in Limitations, not buried.

02 THE FINDING ESTABLISHED

Contradictory public privacy disclosures create a resident-trust exposure that attaches to the City of El Paso — independent of what the platform actually does with data.

Verified by first-party capture of both app-store listings on June 18, 2026; exhibits on file. The same application describes its data behavior three irreconcilable ways in three public places — and two of the three are the developer's own declarations:

Apple App Store — "App Privacy" · developer SuperCity LLC

Data Used to Track You: Usage Data, used to track the user "across apps and websites owned by other companies."
Data Linked to Your Identity: Email Address, Name, and User ID.
Data Not Linked to You: Device ID, Usage Data, Crash Data.

Google Play — "Data safety" · same developer

"This app doesn't collect or share any user data." No data collected. No data shared. (Notes only: encrypted in transit; deletion on request.)

Company spokesperson — on the record, Oct 2025

The company "does not collect or sell resident data to third parties."

The first two are the same company describing the same app on the same day in opposite terms. Apple's listing — SuperCity's own declaration — says it collects email, name, and identifiers and tracks usage across third-party apps. Google's listing — also SuperCity's own declaration — says it collects nothing. Both cannot be true.

That is the exposure. A resident, journalist, or council member who places these side by side does not need to resolve which is accurate — the discrepancy itself is the damage, because it converts a trust question into a credibility question. And because Apple's listing declares collection of named contact information linked to identity, the stakes are not abstract: this is a city-services app whose own disclosure says it gathers residents' email and name.

The core pattern an Institutional Signal Audit looks for: the public reads the authority signal, not the documents. The City's signal said "safe to use." The disclosure layer said something else. The audience receives the gap — and it lands on the institution that vouched for the app.

03 WHY IT MATTERS NOW SUPPORTED

1. The governance instrument is a Letter of Intent, not a data-governance contract. Public reporting describes a "content-only" LOI under which the City pays nothing and the app uses only already-public information. An LOI is the right tool for moving fast — but it gives the City limited contractual leverage over how resident interactions with a city-branded service are handled. There is, on the public record, no visible data-governance addendum specifying collection, retention, sharing, or audit rights. (Full LOI terms are not public; rated Supported, not Established.)

2. The local environment is already primed for data scrutiny. In March 2026 the El Paso City Council moved to reconsider its Flock Safety license-plate-reader camera contract over surveillance and data concerns. A council and press corps already alert to vendor data practices will read a privacy-disclosure contradiction quickly and unforgivingly.

Net: a city-branded AI front door, governed by a thin instrument, in a privacy-sensitive moment, carrying a documented disclosure contradiction. None of this requires bad intent to become a public problem. It only requires someone to notice — and noticing is cheap.

04 THE FIRST FIX PRIORITY ORDER

Reconcile the three disclosures to a single, accurate statement. Whichever is correct, the Apple listing, the Google listing, and the spokesperson language must say the same thing. A days-not-weeks fix that removes the Established-tier exposure outright.

Attach a one-page data-governance addendum to the LOI. Collection, retention, sharing, deletion, and the City's audit rights, in plain terms. Converts the thin instrument into a defensible one without slowing the partnership.

Publish a plain-language resident data statement under the City's name. Turn the liability into a trust asset: the City that got ahead of the question rather than answering it under pressure.

05 WHERE THIS POINTS NEXT STEP

This finding sits at the entry of the governance ladder. The natural next step is a Governance Discovery Sprint — mapping the full SuperCity deployment (data flows, decision points, escalation paths, the LOI's actual terms) to confirm whether this disclosure gap is the only exposure or the most visible one. If the City wants the addendum and resident statement built and a reporting structure designed, that is Governance Design. The Institutional Signal Audit fee credits in full toward either.

06 LIMITATIONS NAMED CONSTRAINTS

Runtime data practice is not adjudicated. This audit establishes that the disclosures contradict one another — verified by first-party capture. It does not establish, by network or technical inspection, what the app in fact transmits or retains at runtime. Resolving which disclosure is accurate requires that technical step (a sprint-level task). The finding does not depend on it: the contradiction between the developer's own two declarations is the exposure, regardless of which is correct.

Disclosures are point-in-time. Both listings were captured June 18, 2026. App-store disclosures can be edited by the developer at any time; a capture is a timestamped snapshot, not a permanent state. Re-verify immediately before any external use.

The LOI's full terms are not public. The governance-gap finding is Supported, not Established, pending review of the actual instrument.

Single-analyst review. No second-coder check. Consistent with a one-week front-door audit; a sprint adds adversarial review.

07 EXHIBITS & SOURCES ON FILE

EXHIBIT A — Google Play "Data safety," SuperCity AI (captured 2026-06-18): "doesn't collect or share any user data."

EXHIBIT B — Apple App Store "App Privacy," developer SuperCity LLC (captured 2026-06-18): Data Used to Track You; Data Linked to You (Email, Name, User ID).

EXHIBIT C — Apple App Store "App Privacy," continued (captured 2026-06-18): Data Not Linked to You (Device ID, Usage Data, Crash Data).

SuperCity Launches with The City of El Paso — SuperCity Newsroom (Sept 2025) — supercity.ai/newsroom/supercity-ai-launches-with-the-city-of-el-paso
AI-powered SuperCity app launches in El Paso — El Paso Inc.
Unveiling of SuperCity App Leads to Concerns in El Paso About Privacy — El Paso Herald Post (Oct 7, 2025)
AI-powered "SuperCity" app partners with El Paso — KVIA (Oct 7, 2025)
SuperCity AI Crafts El Paso AI Apprenticeship Program — GovTech
El Paso City Council may reconsider Flock Safety camera contract — El Paso Matters (Mar 2, 2026)

Prepared as a pilot demonstration of the Institutional Signal Audit method. Findings are referenced to public sources and rated for evidentiary confidence. This document does not allege wrongdoing by any party; it identifies a documented, fixable public-signal exposure and a constructive first step.